Although the imaging technology based on color channels meets the imaging requirements of the human vision system, spectrum information of the scene is lost during the integrating process of the scene light spectrum in the camera sensors. The lost information is the key property for reflecting the physical characteristics of the light source and the scene, and thus it is important to solve the problem of sampling the spectrum information of the scene. Starting from this, many researches on sampling the spectrum information have been performed.
According to different requirements and sampling conditions, the conventional multi-spectral sampling systems comprise a spectrum analyzer, a scanning imaging spectrometer and a single shot imaging spectrometer. All the conventional multi-spectral sampling systems sample the multi-spectral information by compensating the spectral resolution with sacrificing the spatial resolution and time resolution. Thus, how to obtain the scene information having the high spectral resolution, high time solution and high spatial resolution is a problem.
For this, a multi-spectral sampling system which performs a two-path sampling is proposed. Although the multi-spectral information can be sampled well with the above two-path sampling, a precise spatial calibration and spectral calibration cannot be realized.